β-Catenin Stabilization in Skin Fibroblasts Causes Fibrotic Lesions by Preventing Adipocyte Differentiation of the Reticular Dermis  Maria Mastrogiannaki, Beate M. Lichtenberger, Andreas Reimer, Charlotte A. Collins, Ryan R. Driskell, Fiona M. Watt  Journal of Investigative Dermatology  Volume 136, Issue 6, Pages 1130-1142 (June 2016) DOI: 10.1016/j.jid.2016.01.036 Copyright © 2016 The Authors Terms and Conditions

Figure 1 Localization and isolation of Sca1+ and Sca1- dermal fibroblasts. (a–d, n–q) Sections of P2 back skin from PdgfrαEGFP mice showing expression of nuclear EGFP and immunostaining for (a, b) K14 and PDGFRα, (c, d) Sca1, (n, o) Fabp4, and (p, q) Perilipin. (b, d, o, q) are enlargements of selected regions of (a, c, n, p), respectively. Scale bars = 150 μm. (e) Flow cytometry plot showing gating of different dermal subpopulations on the basis of PdgfrαEGFP and Sca1 expression. (f–m) Quantitative real-time PCR analysis of mRNA levels in sorted cell populations: PdgfrαEGFP-/Sca1-, PdgfrαEGFP+/Sca1- and PdgfrαEGFP+/Sca1+. Genes were normalized to the Gapdh gene. Error bars represent standard error of the mean of replicates from four mice. ∗P ≤ 0.05, ∗∗P ≤ 0.005, ∗∗∗P ≤ 0.0005 compared with GFP- cells; #p ≤ 0.05 compared with EGFP+/Sca1- cells. ad, adipocyte; epi, epidermis; der, dermis; EGFP, enhanced green fluorescent protein; GFP, green fluorescent protein; HF, hair follicle; pap, papillary; PDGFRα, platelet-derived growth factor receptor α; ret, reticular. Journal of Investigative Dermatology 2016 136, 1130-1142DOI: (10.1016/j.jid.2016.01.036) Copyright © 2016 The Authors Terms and Conditions

Figure 2 Distinct transcriptional signature of PdgfrαEGFP+/Sca1+ dermal cells. (a) Heat map showing hierarchical clustering (based on entities and samples) of all differentially regulated genes (P < 0.05, change >2-fold) between PdgfrαEGFP+/Sca1+ and PdgfrαEGFP+/Sca1- fibroblasts. (b) Selected genes up-regulated or down-regulated in Sca1+ cells. Values in parentheses represent fold change of each gene. (c) Heat map showing hierarchical clustering (based on entities) of all regulated genes in the Gene Ontology term “Wnt receptor signaling pathway”. (d–k) Quantitative real-time PCR analysis of mRNA levels in sorted cell populations, normalized to Gapdh gene expression. Error bars represent standard error of the mean of replicates from four mice. ∗P ≤ 0.05, ∗∗P ≤ 0.005, ∗∗∗P ≤ 0.0005 compared with GFP- cells; #P ≤ 0.05 compared with GFP+/Sca1- cells. (l, m) Immunofluorescent staining of neonatal skin with an antibody detecting β-catenin. Red arrowheads show β-catenin+ fibroblasts in the reticular dermis. 4′, 6-diamidino-2-phenylindole labels nuclei. Scale bar = 200 μm. (n) Section of P1 back skin immunostained for Tcf3/4 (red) and Lef1 (green). White arrowheads indicate double-labeled cells. Dashed lines demarcate epidermal-dermal boundary. Scale bar = 100 μm. (o–r) Higher-magnification images of the boxed areas in (m, n), showing upper (o, q) and lower (p, r) dermis. BMP, bone morphogenic protein; DAPI, 4′, 6-diamidino-2-phenylindole; GO, Gene Ontology; PDGF, platelet-derived growth factor. Journal of Investigative Dermatology 2016 136, 1130-1142DOI: (10.1016/j.jid.2016.01.036) Copyright © 2016 The Authors Terms and Conditions

Figure 3 β-Catenin stabilization in dermal fibroblasts via PdgfrαCreERT2. (a) Schematic illustration of experimental strategy. (b) Recombination efficiency of PDGFRαCreERT2. Bar graph showing percentage of tdTomato+ cells within different fibroblast subsets after gating for ITGA6-/LIN-/CD31- cells 2 days after 4-OHT–mediated recombination. Data are reported as mean ± standard error of the mean of triplicate samples in a representative experiment (n = 2–3 independent experiments). (c, d) tdTomato expression in 4-OHT–treated dorsal skin from (c) PDGFRαCreERT2 × tdTomato-LSL × Ctnnb1 Exon3+/+ (control) and (d) PDGFRαCreERT2 × tdTomato-LSL × Ctnnb1 Exon3Flox/+ (cΔex3) mice at P4. (e–h) Strong β-catenin staining is detected in tdTomato+ fibroblasts of mice with activated Wnt/β-catenin signaling. Note that the PDGFRαCreERT2 transgene is also active in cells of the dermal papilla (white arrowheads in g). Red arrow heads in (h) depict β-catenin+ cells in the dermal papilla. (i–p) Paraffin sections of back skin of control and mutant (cΔex3) littermates stained for hematoxylin and eosin at (i, j) P4, (k, l) first telogen (P18), (m, n) anagen (P35), and (o, p) second telogen (P56). Boxed areas in (k, l) are shown as higher-magnification inserts. Scale bars = 200 μm. (q, r) Hair follicle length measured in (q) male and (r) female cΔex3 and control mice at P4, P11, P18, P24, P28, P35, P42, and P56 (n = 2–8). Data points are reported as mean ± standard error of the mean. (s) Quantification of proliferating, phosphohistone H3+ cells in the outer root sheath of hair follicles in immunostained skin sections (n = 3 biological samples; ≥5 hair follicles per biological sample were scored). 4-OHT, 4-hydroxy-tamoxifen; ctr, control; Dapi, 4′, 6-diamidino-2-phenylindole; K14, keratin 14. Journal of Investigative Dermatology 2016 136, 1130-1142DOI: (10.1016/j.jid.2016.01.036) Copyright © 2016 The Authors Terms and Conditions

Figure 4 Induction of dermal fibrosis and proliferation by β-catenin stabilization. (a, b) Herovici’s staining differentiates between immature collagen fibers (blue) versus mature collagen (pink). Note fibrotic region with mature extracellular matrix in (b) (boxed). (c) Relative expression of adipocyte differentiation genes in flow-sorted tdtomato+ fibroblasts. Data represent mean ± standard error of the mean (n = 3). ∗P ≤ 0.05. (d–i) Immunostained paraffin sections of 35-day-old mutant and control skin. (j, k) Whole-mount thick sections stained with LipidTox (green) with DAPI counterstain (blue). (l, m) Quantification in paraffin sections of P35 skin of (l) adipocytes and (m) EdU+ fibroblasts in the adipocyte layer and interfollicular dermis (including papillary fibroblasts, dermal sheath, and reticular fibroblasts). The boundary between papillary/reticular and adipocyte layers was defined by the interface between the bottom of the reticular layer and the upper layer of differentiated adipocytes. Data points represent mean ± standard error of the mean; n ≥ 5 for each group. ∗P ≤ 0.05; ∗∗P ≤ 0.005; ∗∗∗P ≤ 0.0005. (n, o) Skin labeled with EdU (green) for 4 hours before harvesting, costained for Pdgfrα (red), keratin 14 (white), and DAPI (blue). White arrowheads indicate EdU+ fibroblasts within the adipose layer. Inserts are higher-magnification images of the boxed areas. Scale bars = 200 μm. ctr, control; DAPI, 4′, 6-diamidino-2-phenylindole; K14, keratin 14. Journal of Investigative Dermatology 2016 136, 1130-1142DOI: (10.1016/j.jid.2016.01.036) Copyright © 2016 The Authors Terms and Conditions

Figure 5 β-Catenin stabilization in the lower dermis via Dlk1CreERT2. (a) Schematic illustration of experimental strategy. (b, c) Quantification of tdTomato+ cells in E18.5 dermal cell suspensions (ITGA6-/LIN-/CD31-) 2 days after 4-OHT injection. (b) Percentage of tdTomato+ cells in each fibroblast subpopulation. (c) Contribution of each fibroblast subpopulation to tdTomato+ fibroblasts. Data are presented as mean ± standard error of the mean of quadruplicate samples in a representative experiment (n = 3 independent experiments). (d, e) tdTomato-expressing cells in P56 back skin of Dlk1CreERT2 × tdTomato-LSL × Ctnnb1 Exon3+/+ (control) and Dlk1CreERT2 × tdTomato-LSL × Ctnnb1 Exon3Flox/+ (Dlk1LTΔex3) mice. Asterisks in (d) indicate terminally differentiated adipocytes. (f, g) Hematoxylin and eosin and (h, i) Herovici’s staining of paraffin sections of P56 control and Dlk1LTΔex3 mouse back skin. Arrows in (g) and (i) indicate fibrotic regions. (j–m) P56 control and Dlk1LTΔex3 back skin labeled for β-catenin or caveolin-1 and keratin 14, counterstained with DAPI. Vertical lines indicate thickness of adipocyte layer. Dashed line in (k) demarcates fibrotic tissue (FT). (n, o) Quantification of fibrotic area shown as percentage of the adipocyte layer (n) and number of adipocytes (o). (p) Quantification of EdU-labeled fibroblasts in the interfollicular dermis (including papillary fibroblasts, dermal sheath, and reticular fibroblasts) and adipose layer (defined as in Figure 4m) after a 2-hour EdU pulse. Data are presented as mean ± standard error of the mean of triplicate samples in a representative experiment (n = 3–5). ∗P ≤ 0.05; ∗∗P ≤ 0.005. (q, r) EdU staining of sections shown in (d) and (e). Scale bars = 200 μm. 4-OHT, 4-hydroxy-tamoxifen; AL, adipocyte layer; ctr, control; DAPI, 4′, 6-diamidino-2-phenylindole; K14, keratin 14. Journal of Investigative Dermatology 2016 136, 1130-1142DOI: (10.1016/j.jid.2016.01.036) Copyright © 2016 The Authors Terms and Conditions

Figure 6 Effect of β-catenin stabilization on adipocyte differentiation in cultured Sca1+ fibroblasts. (a) Schematic illustration of experimental strategy. (b–d) Sca1+/tdTomato+ targeted fibroblasts from dermal cell suspensions of PDGFRαCreERT2 × tdTomato-LSL × Ctnnb1 Exon3+/+ (control) and PDGFRαCreERT2 × tdTomato-LSL × Ctnnb1 Exon3Flox/+ (cΔex3) littermates were isolated by flow cytometry. (b, c) Gating out α6 integrin-positive cells (ITGA6; keratinocytes) (b) and positive selection for Sca1+/tdTomato+ cells (c). (d) Recombination efficiency (percentage tdTomato+ cells) in Cd24+/Sca1- and Cd24-/Sca1+ cells. Data are presented as mean ± standard error of the mean of triplicate samples in a representative experiment (n = 3 independent experiments). (e) Examples of individual colonies formed by Sca1+ fibroblasts stained for LipidTox (green), tdTomato (red), and DAPI (blue). Scale bars = 10 μm. (f, g) Percentage of colonies containing 1, 2, 3–5, or more than 5 cells in standard (f) and adipogenic (g) medium. (h, i) Total LipidTox fluorescence (lipid intensity) per clone in control and mutant (cΔex3) cultures grown in standard (h) or adipogenic (i) medium. (f–i) n = 2–3 biological replicates and 2 technical replicates. ∗P ≤ 0.05, ∗∗P ≤ 0.01, ∗∗∗P ≤ 0.005. (f, g) two-way analysis of variance with a Bonferroni posttest. (h, i) 25% confidence intervals are shown. 4-OHT, 4-hydroxy-tamoxifen; ctr, control DAPI, 4′, 6-diamidino-2-phenylindole; FSC, forward scatter; ns, difference not significant. Journal of Investigative Dermatology 2016 136, 1130-1142DOI: (10.1016/j.jid.2016.01.036) Copyright © 2016 The Authors Terms and Conditions